PAPR Reduction of OFDM Signals With Minimized EVM via Semidefinite Relaxation

In this paper, we exploit convex optimization techniques for reducing the peak-to-average power ratio (PAPR) of orthogonal frequency-division multiplexing (OFDM) signals. Our optimization strategy is to minimize error vector magnitude (EVM) with constraints of PAPR and free carrier power overhead. Then, a nonconvex quadratic program is obtained, which is computationally difficult. By utilizing the semidefinite relaxation (SDR) technique, the optimization model is relaxed to a semidefinite program, which can be efficiently solved using polynomial time. On top of this, we use the rank minimization technique to improve the proposed model. Simulation results demonstrate that the optimized OFDM signals have quasi-constant PAPR values and are not sensitive to the nonlinear effects of the power amplifier in comparison with other approaches.